J. theor. Biol. (1989) 136, 337-356 Reviving the Superorganism DAVID SLOAN WILSON AND ELLIOTr SOBER Department of Biological Sciences, State University of New York, Binghamton, New York 13901, U.S.A. and Department of Philosophy, 5185 Helen C. White Hall, University of Wisconsin, Madison, Wisconsin 53706, U.S.A. (Received 25 July 1988, and accepted in revised form 18 October 1988) Individuals become functionally organized to survive and reproduce in their environ- ments by the process of natural selection. The question of whether larger units such as groups and communities can possess similar properties of functional organization, and therefore be regarded as "'superorganisms", has a long history in biological thought. Modern evolutionary biology has rejected the concept of superorganisms, explaining virtually all adaptations at the individual or gene level. We criticize the modern literature on three counts. First, individual selection in its strong form is founded on a logical contradiction, in which genes-in-individuals are treated differently than individuals-in-groups or species-in-communities. Imposing con- sistency clea~rly shows that groups and communities can be organisms in the same sense that individuals are. Furthermore, superorganisms are more than just a theoreti- cal possibility and actually exist in nature. Second, the view that genes are the "ultimate" unit of selection is irrelevant to the question of functional organization. Third, modern evolutionary biology includes numerous conceptual frameworks for analyzing evolution in structured populations. These frameworks should be regarded as different ways of analyzing a common process which, to be correct, must converge on the same conclusions. Unfortunately, evolutionists frequently regard them as competing theories that invoke different mechanisms, such that if one is "right" the others must be "wrong". The problem of multiple frameworks is aggravated by the fact that major terms, such as "'units of selection", are defined differently within each framework, yet many evolutionists who use one framework to argue against another assume shared meanings. We suggest that focusing on the concept of organism will help dispell this fog of semantic confusion, allowing all frameworks to converge on the same conclusions regarding units of functional organization. Introduction The idea that single-species groups, multi-species communities, and human societies can possess the properties of single organisms is a frequently recurring theme in human thought. Within biology, social insect colonies have been regarded in this fashion for centuries. More recently, Clements (1916) compared the succession of plant species that culminates in a forest, to the growth and development of a single organism. Odum (1969) compared ecosystem processes such as nutrient cycling to the physiological processes of individuals. Wynne-Edwards (1962, 1986) proposes that populations evolve to avoid extinction, just as individuals evolve to avoid their own deaths. Margulis (1970, 1981) proposes that eucaryotic cells are in fact tightly integrated communities of bacteria. Lovelock (1979) even portrays the entire earth as a kind of organism that regulates its atmosphere to be conductive for life. 337 0022-5193/89/030337 + 21 $03.00/0 O 1989 Academic Press Limited 338 D.S. WILSON AND E. SOBER These theories have a certain magnetic appeal, as their sheer number and popular- ity among lay-people attests. Nevertheless, the superorganism concept has not fared well within evolutionary biology, which is the one discipline best qualified to judge it. Many evolutionists subscribe to a concept known as individual selection, which holds that groups, communities and human societies are only collections of organ- isms without themselves having the properties of organisms. Individual selectionists not only regard the existence of superorganisms as a dead issue, but they also regard its death as one of their greatest achievements (Williams, 1966; Ghiselin, 1974; Dawkins 1976, 1982; Maynard Smith, 1987a, b). The demise of the superorganism is perhaps best illustrated by the following quotes from modern social insect biologists. "Despite the logical force of arguments against group (or colony) selection, and the invention of tidy explanations for collaboration in individual terms, the supraorganism (colony-level selection) still haunts evolutionary discussions of insect sociality." (West- Eberhard, 1978, p. 12) "I know of no observation on honeybee biology which unequivocally demonstrates the action of colony-level selection working at the expense of individual interests." (Seeley, 1985, p. 7) Thus, according to these authors, even something as organism-like as a bee colony must be understood in individualistic terms. It is true that many superorganism theories are best regarded as poetic metaphors in scientific guise. To regard the earth's atmosphere as a fragile goddess that needs protection is far more compelling than to regard it as a complex system of gases that needs management. Unfortunately, the metaphor creates an illusion of func- tional organization where none exists, and presents a misleading image of nature as inherently benign and cooperative. The superorganism concept fails as a grandiose theory of nature, and its death in this form is indeed a triumph of modern evolution- ary biology. Against this background, reviving the superorganism concept might seem like bringing back Dr Frankenstein's well-intentioned monster. Nevertheless, at least five compelling reasons exist for constructing a clearly formulated scientific theory of superorganisms. (i) Individual selection in its strong form is founded on a logical contradiction. Imposing consistency clearly shows that collections of individuals can become functionally organized by natural selection, in exactly the same way that individuals themselves become functionally organized. (ii) Conditions for the evolution of superorganisms are not overly restrictive. Superorganisms are more than just a theoretical possibility; they actually exist in nature. (iii) The theory predicts that only some groups and communities qualify as superorganisms. Others are mere collections of individuals or complex mixtures of adaptation at more than one level. Thus, a legitimate theory of superorganisms is not grandiose, and can avoid the excesses of the past. (iv) The strong form of individual selection itself is a metaphor that creates a misleading picture of nature as inherently exploitative and competitive. The notion REVIVING THE SUPERORGANISM 339 that all evolved adaptations are brands of self-interest must be substantially modified if individuals sometimes function as "alleles" in group and community "organisms". A legitimate theory of superorganisms might therefore help correct certain excesses of the present. (v) Major terms surrounding this subject, such as "individual selection", "group selection", etc. have acquired multiple and conflicting meanings. Several conceptual frameworks exist that masquerade as competing theories, but which actually are alternative ways of analyzing a common process of evolution in structured popula- tions. To be correct, each framework must therefore reach the same conclusion about units of functional organization. This convergence can be facilitated by focusing on the concept of organism within each framework. Semantics The term "individual" and "organism" have a variety of meanings in evolutionary biology, and before proceeding it is important to clarify our own usage. Hull (1980) defines individuals as "spatio-temporally localized entities that have reasonably sharp beginnings and endings in time". Notice that this definition make~ no reference to functional organization; an atom, a gene, a single creature (such as an insect) and an entire species all are examples of individuals. By contrast, an organism can be defined as "a form of life composed of mutually dependent parts that maintain various vital processes" (Random House dictionary, unabridged edition). Thus, the hallmark of an organism is functional organization. An atom is not an organism, genes (and the organs they code for) are only parts of an organism, and species are frequently only collections of organisms. Because single creatures are, in most respects, both individuals and organisms, biologists informally use the two words interchangeably. The term ~'individual selection", for example, refers to single creatures, and not to the idea encompassed by Hull's broader definition. We will abide by the colloquial synonymy and refer to single creatures as both individuals and organisms. When more formal definitions are required, we will be careful to identify them as such. We define a superorganism as a collection of single creatures that together possess the functional organization implicit in the formal definition of organism. Just as genes and organs do not qualify as organisms, the single creatures that make up a superorganism also may not qualify as organisms in the formal sense of the word. We will continue to refer to them as individual organisms, however, to conform to colloquial usage. Individuals as Organisms, and the Concept of Heritability Individuals acquire the exquisite functional organization that justifies their status as organisms by the process of natural selection. As Darwin realized, natural selection requires heritable variation. Individuals must differ in their properties, and those differences also must exist to some degree in their offspring. If so, then